Brain Struct Funct DOI 10.1007/s00429-015-1071-2
ORIGINAL ARTICLE
Age-related differences in task goal processing strategies during action cascading Ann-Kathrin Stock1 • Krutika Gohil1 • Christian Beste1
Received: 14 January 2015 / Accepted: 22 May 2015 Ó Springer-Verlag Berlin Heidelberg 2015
Abstract We are often faced with situations requiring the execution of a coordinated cascade of different actions to achieve a goal, but we can apply different strategies to do so. Until now, these different action cascading strategies have, however, not been examined with respect to possible effects of aging. We tackled this question in a systems neurophysiological study using EEG and source localization in healthy older adults and employing mathematical constraints to determine the strategy applied. The results suggest that older adults seem to apply a less efficient strategy when cascading different actions. Compared to younger adults, older adults seem to struggle to hierarchically organize their actions, which leads to an inefficient and more parallel processing of different task goals. On a systems level, the observed deficit is most likely due to an altered processing of task goals at the response selection level (P3 ERP) and related to changes of neural processes in the temporo-parietal junction. Keywords Executive control Action cascading Older adults EEG Source localization Temporo-parietal junction
A.-K. Stock and K. Gohil contributed equally. & Christian Beste
[email protected] Ann-Kathrin Stock
[email protected] 1
Cognitive Neurophysiology, Department of Child and Adolescent Psychiatry, Faculty of Medicine of the TU Dresden, Schubertstrasse 42, 01307 Dresden, Germany
Introduction Action control is important for daily activities. It is well known that older adults have difficulties in action control when two tasks have to be coordinated at almost the same time (e.g., Chmielewski et al. 2014; Hartley et al. 2011; Verhaeghen et al. 2003; Verhaeghen and Cerella 2002). Instead of being required to simultaneously execute several actions at exactly the same time, we are, however, much more often faced with situations requiring the execution of a coordinated cascade of different actions to achieve a goal. For example, when driving a car, you may be required to first stop the car in front of stop sign even though your navigation system instructs you to immediately turn right after this sign. Action cascading processes can be examined using a stop-change task that requires rapid responses in reaction to certain stimuli. Occasionally, these reactions have to be interrupted (stopped) upon the delivery of a STOP stimulus, and an alternative response has to be executed upon the presentation of another stimulus (CHANGE stimulus). The time to change is indicated by the reaction time (RT) to this CHANGE stimulus. The CHANGE stimulus may either be presented at the same time with the STOP stimulus or after a certain time interval. In conditions where STOP and CHANGE stimuli are presented at the same time, people can choose to try to process both stimuli at once (overlapping/parallel strategy), or prefer processing of the stimuli in a step-by-step fashion (serial strategy), i.e., finishing the stop process before turning to the change process. It has been shown that the ‘‘overlapping strategy’’ is related to inefficient unfolding of behavior, i.e., longer reaction times on CHANGE stimuli, because the stopping and changing processes have to share limited response selection capacity (Miller et al. 2009; Mu¨ckschel et al. 2014; Verbruggen et al. 2008). Until now,
123
Brain Struct Funct
such action cascading processes (Mu¨ckschel et al. 2014) have not been examined regarding possible effects of aging. It is, however, possible that action cascading processes are differentially modulated in older and younger adults for a number of reasons: It has been shown that the striatum plays an important role for action cascading processes. In particular, it has been shown that dysfunctions of the striatum lead to a less efficient strategy when cascading different actions (e.g., Beste et al. Beste et al. 2014a, b; Beste and Saft 2015; Ness and Beste 2013; Yildiz et al. 2014). Moreover, previous electrophysiological studies reflect the roles of the anterior cingulate cortex (ACC) and temporo-parietal junction (TPJ) in action cascading efficiency (Mu¨ckschel et al. 2014). Other results suggest that the efficiency of action cascading processes strongly depends on the functioning of the dopaminergic system (Stock et al. 2014a, b) and factors interacting with the dopaminergic system (Beste et al. 2014a, b). Fronto-striatal structures, ACC, TPJ and the dopaminergic system are well known to show dysfunctional changes during aging (e.g., Buckner 2004; Klostermann et al. 2012; Nieoullon and Coquerel 2003). As these neurobiological and functional neuroanatomical systems are central for efficient action cascading and are somewhat dysfunctional in the older adults, we hypothesize that older adults show a less efficient strategy/mode than younger adults when cascading different actions. In the current study, we, therefore, examine the hypothesis that older adults show a less efficient strategy/mode than younger adults when cascading different actions, as measured on a behavioral and neurophysiological level using event-related potentials (ERPs) and source localization techniques (sLORETA). We employ a stop-change paradigm (e.g. Mu¨ckschel et al. 2014; Verbruggen et al. 2008) that makes it possible to estimate the employed strategy and efficiency of action cascading on the basis of mathematical constraints. Previous results suggest that the critical mechanisms that determine the efficacy of action cascading processes occur at the level of response selection (Mu¨ckschel et al. 2014). The P3 ERP, which has been suggested to reflect response selection processes occurring between stimulus processing and the overt motor response (e.g. Verleger et al. 2005), is modulated by the applied action cascading strategy (Mu¨ckschel et al. 2014). It has been shown that subjects show strong differences in the modulation of the P3 between conditions that either enforce a serial activation of actions vs. conditions that leave the subjects with a choice as to whether they process their actions in a step-by-step (serial) fashion or simultaneously (Mu¨ckschel et al. 2014). Due to the above-mentioned age-related changes, we expect older adults to show a less efficient strategy during action cascading processes. That is, we expect older adults
123
to show a more parallel mode of task goal activation during action cascading. On a neurophysiological level, we expect older adults to show age-related changes in the P3, but not in attentional ERP components (the P1 and N1) which do not seem to play a major role in different strategies applied during action cascading (Mu¨ckschel et al. 2014). On a systems level, we expect that these changes are reflecting a fronto-parietal network comparable to the multiple demand (MD) system (Duncan 2010), since fronto-parietal regions have been shown to underlie differences in the applied processing strategies (Mu¨ckschel et al. 2014).
Materials and methods Sample Fifteen older adults were recruited for the study, but two of them had to be excluded due to poor EEG data quality. The remaining n = 13 older adults (6 females, 7 males) who participated in the study were 64.53 ± 3.3 years old and stated to be right-handed. A sample of n = 17 younger (24.05 ± 2.88 years of age) right-handed participants (11 females, 6 males) was also recruited. There was no difference in the level of depressive symptoms between younger adults (3.6 ± 4.6) and older adults (6 ± 4.1) as assessed using the Beck Depression Inventory (BDI; Beck et al. 1961) (p [ 0.3). The same was found for performance in the Mini Mental Status Examination (MMSE) (young adults: 29 ± 0.2; older adults: 29 ± 1.2; p [ 0.6). There was also no difference in the years of education (young adults: 12.2 ± 2.2; older adults: 11.4 ± 3.5; p [ 0.6). Each participant gave written informed consent before beginning the experiment. After completing the experiment, each of them was reimbursed with 20 €. The study was approved by the ethics committee of the Medical faculty of the TU Dresden. Experimental paradigm As shown in Fig. 1, a modified version of the stop-change paradigm introduced by Verbruggen et al. (2008) was used in this experiment. The experiment was conducted in a sound-attenuated room where each subject was comfortably seated at a distance of 56.5 cm from a 21 inch computer monitor. To record the participants’ responses, a custom-made keyboard with four different keys was placed in front of them. The presentation of experimental stimuli, the recording of behavioral responses and sending triggers to the computer recording the EEG were attained by using ‘‘Presentation’’ software (Neurobehavioral Systems, Inc.).
Brain Struct Funct
Fig. 1 Schematic illustration of the stop-change paradigm. GO trials end after the first response to the GO stimulus (bold). SC trials end after the first response to the CHANGE stimulus (bold). SC require stopping of the GO response upon presentation of the STOP signal.
After stopping, an alternative response to the CHANGE stimulus has to be executed. For this, the new yellow reference line has to be taken into account. The stimulus onset asynchrony (SOA) between the onset of the STOP and CHANGE stimuli was set to either 0 or 300 ms
The experimental task took the participants about 25 min to complete and comprised 864 trials (divided into six blocks). Out of these trials, two thirds were GO trials and the remaining third were stop-change (SC) trials. All trials were presented in a pseudo-randomized order to avoid preparatory effects in the motor system. Stimuli were presented against a black background. The task array consisted of 4 vertically arranged, white-bordered circles separated by 3 white horizontal bars (reference lines), which were enclosed in a white-bordered rectangle (as shown in Fig. 1). The circles and reference lines were separated by 0.5 cm. Each trial began with this empty array and after 250 ms; one of the four circles was always filled in with white color. In GO trails, this white circle became the target and participants were asked to respond with the right hand by pressing one of two keys on their right hand side: In case the target was located above the middle reference line, participants had to respond with their right middle finger and had to respond with their right index finger if the target was located below that line. If participants did not respond within 1000 ms after the onset of the target, a speed-up sign (containing the German word ‘‘Schneller!’’) (which translates to ‘‘Faster!’’) was presented above the stimulus array until the trial was ended by a button press. Stop-change (SC) trials also began with the empty array followed by the GO stimulus. However, after a variable
stop signal delay (SSD), the GO stimulus was followed by a STOP stimulus (the border of the rectangle turned from white to red (see Fig. 1), instructing the participants to stop (interrupt) the already initiated right hand GO response. The stop stimulus was then followed by a CHANGE stimulus. There were two CHANGE conditions: In the first condition, there was no delay between the STOP and the CHANGE stimuli (i.e., a stop-change delay of 0 ms, called SCD0) so that stop and change stimuli were presented simultaneously. In the second condition, there was a stopchange delay of 300 ms (SCD300) so that the CHANGE stimulus was presented 300 ms after the onset of the STOP stimulus. The change stimuli were yellow bars, which remained on the screen until the participant responded by pressing one of the response keys. In each SC trial, one of the three horizontal lines would turn into a thick yellow bar, thus becoming the new reference line that needed to be attended. The participants were asked to spatially relate the target (white circle) to the new reference line. In case the target was located above the yellow reference line, participants had to respond with their left middle finger and when the target was located below the reference line, then participants had to respond with their left index finger. In case participants did not respond within 2000 ms after the onset of the CHANGE stimulus, the speed-up sign was presented above the stimulus array until the trial was ended by a button press.
123
Brain Struct Funct
The SSD described above was initially set to 250 ms and dynamically adjusted to the performance by means of a staircase algorithm (Verbruggen et al. 2008). When the participant did not make any mistakes during an SC trial (i.e., did not respond before the presentation of the STOP stimulus and correctly responded to the CHANGE stimulus), the SSD for the following SC trial was increased by 50 ms. In case of any incorrect response, the SSD was decreased by 50 ms. Hence, the staircase yielded a 50 % probability of successful inhibition upon stop signal presentation. To keep the trial duration within reasonable limits, SSD variation was restricted to a range from 50 to 1000 ms. Estimation of the used strategy and efficiency of action cascading The strategy and the efficiency of cascading these different actions were estimated using a mathematical model. As described above, the paradigm introduces two different SCD intervals. These intervals are important for the estimation of the applied strategy, given that there are capacity limitations applying to response selection processes (e.g., Mu¨ckschel et al. 2014; Stock et al. 2014a; Verbruggen et al. 2008). The SCD0 condition provides the participants with a ‘‘choice’’ of how to process the task requirements put forward by the STOP and CHANGE stimuli. Because response selection depends on a restricted resource, the applied strategy affects reaction times (RTs): if participants choose to process STOP- and CHANGE-associated task goals at the same time (i.e. in parallel), RTs increase because these processes have to share a limited capacity. In the SCD0, the condition, the participants can, however, also choose a strategy in which they process STOP- and CHANGE-associated task goals in a step-by-step (i.e., serial) manner. If participants choose such a serial strategy, the STOP and CHANGE processes do not have to share a limited capacity because the STOP process is finished before the CHANGE process starts. This leads to shorter RTs, as compared to the parallel strategy where STOP- and CHANGE-associated task goals are processed at the same time. Critically, the SOA of 300 ms enforces a serial processing of the STOP- and CHANGE-related processes in the SCD300 condition because the STOP process has usually finished when the CHANGE stimulus is presented 300 ms later. Therefore, comparing this SCD300 ‘‘baseline’’ to the SCD0 condition makes it possible to estimate which processing strategy has been used in the SCD0 condition. If a parallel processing strategy has been used in the SCD0 condition, RTs are much longer in the SCD0 than in the SCD300 condition. If a more serial processing strategy has been used in the SCD0 condition, RTs should
123
be less prolonged and rather comparable to those of the SCD300 condition. The ratio of RT differences in the SCD0 and SCD300 conditions, therefore, gives an estimate about the strategy applied during action cascading: slope ¼ ðRTSCD0 RTSCD300 Þ=ðSCD0 SCD300 Þ This slope value was individually calculated for each participant. It becomes steeper the more RTSCD0 and RTSCD300 differ. When the STOP process has not finished at the time the CHANGE process is initiated (parallel processing strategy), the slope value becomes larger, indicating that action cascading is less efficient (Schwarz and Ischebeck 2001; Verbruggen et al. 2008). If it has finished (serial processing strategy), the slope approaches zero, which indicates that action cascading becomes more efficient (Schwarz and Ischebeck, 2001; Verbruggen et al. 2008). Obtaining a mean slope value between 0 and -1 hence suggests that some (but not all) of the CHANGE response processes were initiated before the termination of the inhibitory process stopping the GO response. Therefore, the slope of the SCD RT function is flatter in case of a more efficient processing strategy than in the case of a less efficient processing strategy. EEG recording and analysis High-density EEG recording was acquired using a QuickAmp amplifier (Brain Products, Inc.) with 60 Ag–AgCl electrodes at standard scalp positions in an equidistant electrode setup. The reference electrode was located at electrode Fpz. The data were recorded with 1 kHz and then down-sampled offline to 256 Hz. All electrode impedances were kept below 5 kX. Afterwards, an IIR band-pass filter ranging from 0.5 to 20 Hz with a notch at 50 Hz was applied. A manual inspection of the data was performed to remove technical artifacts. Then, to correct the periodically recurring artifacts such as pulse, eye blinks or saccade artifacts, an independent component analysis (ICA) was applied using the infomax algorithm. Afterwards, one more manual raw data inspection was applied to remove any remaining artifacts. After the removal of artifacts, the EEG data was segmented according to the two different SCD conditions. The segmentation was performed in relation to the occurrence of the stop signal (see also: Mu¨ckschel et al. 2014). After the data were epoched, an automated artifact rejection was applied. The rejection criteria included a voltage of more than 150 lV/ms, a value difference of more than 150 lV in a 250-ms interval, or activity below 0.1 lV in a 100 ms interval. To eliminate the reference potential from the data, a current source density (CSD) transformation was run (Perrin et al. 1989). In addition to removing the reference potential, the CSD also serves as a spatial filter (Nunez and Pilgreen 1991) which helps to
Brain Struct Funct
identify the electrodes that best reflect activity related to cognitive processes. Finally, a baseline correction was made for the time window from -900 to -700 ms. Next, the P1, N1, and P3 event-related potentials (ERPs) were quantified. Electrodes were chosen on the basis of the scalp topographies. The visual P1 and N1 were quantified at electrodes P7 and P8 (P1: 90– 140 ms and N1: 190–300 ms post-stimulus, respectively), and the P3 was quantified at Cz (SCD0: 290–410 ms and SCD300: 310–460 ms). All ERP components were quantified relative to the pre-stimulus baseline. For all components, we quantified peak amplitude and latency on the single-subject level. Statistics Mixed effects analyses of variance (ANOVAs) were used to separately analyze behavioral and ERP data. The factors ‘‘condition’’ (SCD0 and SCD300) and ‘‘electrode’’ (only for ERP data) were used as within-subject factors. The factor ‘‘group’’ (older adults vs. younger adults) was used as a between-subjects factor. The degrees of freedom were adjusted accordingly using Greenhouse-Geisser correction. All post hoc tests were Bonferroni-corrected. Kolmogorov–Smirnov tests indicated that all variables used for the analysis were normally distributed (all z \ 0.5; p [ 0.4). The standard error of the mean (SEM) was used as a measure of variability.
Results Behavioral data The analysis of RTs on GO trials revealed differences between the age groups [F(1, 28) = 46.97; p \ 0.001; g2 = 0.627]. RTs of the older group were longer (782 ms ± 132) than those of the younger adults (511 ms ± 83). A mixed effects ANOVA using the withinsubject factor SCD interval and the between-subject factor group revealed a main effect of SCD interval [F(1, 28) = 432.80; p \ 0.001; g2 = 0.939], indicating that RTs were longer in SCD0 trials (883 ms ± 28) than in SCD300 trials (686 ms ± 28). Also, there was a main effect of group [F(1, 28) = 34.83; p \ 0.001; g2 = 0.554] showing that RTs were generally longer in the older adults (951 ms ± 42) than in the younger adults (619 ms ± 37). Moreover, there was an interaction of SCD interval 9 group [F(1, 28) = 9.24; p = 0.005; g2 = 0.248], indicating differential effects of age on RTs in the two SCD conditions. Post hoc independent samples t tests were used to examine the interaction in more detail. They revealed that group differences were more pronounced in the SCD0 condition [older adults: 1063 ms ± 153; younger adults: 703 ms ± 90;
t(28) = 9.34; p \ 0.001] than in the SCD300 condition [older adults: 838 ms ± 205; younger adults: 535 ms ± 100; t(28) = 5.88; p \ 0.001]. Calculating the slope for the SCD RT function (Verbruggen et al. 2008) revealed that the two groups differed from each other [t(28) = -3.049; p = 0.005]. The older adults showed a steeper slope (-0.75 ± 0.18) than the younger adults (-0.56 ± 0.16). This indicates that the degree of difference between the two SCD conditions is larger in the older adults than in the younger adults, suggesting that older adults either employ a less efficient and more parallel response selection strategy than younger adults or that they are processing action cascading serially but their competency to inhibit the go response has declined with aging. The SSRT differed between groups (t = 4.559; p \ 0.001). Matching previous findings (e.g. Coxon et al. 2014, 2012; Sebastian et al. 2013; Williams et al. 1999), the SSRT value was larger in the old (565 ms ± 261) than in the young (244 ms ± 83) group. On SC trials, the accuracy of the STOP response cannot differ due to the applied staircase procedure. Another consequence from the staircase procedure was the main effect of SCD interval found in the number of correct responses to the CHANGE stimulus [F(1, 27) = 489.73; p \ 0.001; g2 = 0.946]. It showed that accuracy was higher in the SCD300 condition (118 ± 3.3) than in the SCD0 condition (76 ± 2.3). There was no interaction of SCD interval 9 group [F(1, 28) = 0.94; p [ 0.5]. Summarizing the behavioral data, we found that the older adults has a steeper slope as compared to the younger adults, which indicates that the older adults apply a more parallel mode of task goal activation during action cascading than younger adults. A speed-accuracy trade-off can be ruled out because the accuracy did not show a differential modulation across SCD conditions and groups. Neurophysiological data P1 and N1 The P1 and N1 ERPs are shown in Fig. 2. Electrodes P7 and P8 are located in the center of the scalp positivities (P1) and the negativities (N1). P1 ERPs were analyzed in mixed effects ANOVA using the factors SCD interval and electrode as within-subject factors and group as a between-subject factor. For P1 amplitudes, there were no main effects of SCD interval, electrode, or group (all F \ 0.9; p [ 0.5). Moreover, there were no interactions of SCD interval 9 group, of electrode 9 group, of SCD interval 9 electrode, or of SCD interval 9 electrode 9 group (all F \ 0.9; p [ 0.5). A similar pattern of results was found for the P1 latency. There was no interaction of SCD interval 9 group, of electrode 9 group, of SCD interval 9 electrode, or of SCD interval 9 electrode 9 group (all F \ 0.9;
123
Brain Struct Funct
Fig. 2 Event-related current source densities for the P1 and N1 in the older adults’ group (left) and the younger adults’ group (right) at electrodes P7 and P8. Time point 0 denotes the time point of the STOP stimulus onset. In the SCD0 condition, the CHANGE stimulus is presented simultaneously while in the SCD300 condition, the onset of the CHANGE stimulus does not occur until 300 ms after the onset
of the STOP stimulus. The color of the ERPs denotes the different SCD conditions. The scalp topography plots of the P1 and N1 peaks show the distribution of the positive and negative potentials across the back of the skull. In the scalp topography plots, cold colors indicate negativity and warm colors denote positivity
p [ 0.5). Also, there were no main effects of group or SCD interval (all F \ 0.9; p [ 0.5). However, results revealed a main effect of electrode [F(1,28) = 10.43; p = 0.003; g2 = 0.271], which showed that the P1 latency was longer at electrode P7 (128.69 ms ± 2.17) than at electrode P8 (122.12 ms ± 2.47). Similar to the P1, N1 ERPs were analyzed in mixed effects ANOVA using the factors SCD interval and electrode as within-subject factors and group as a between-subject factor. For the N1 amplitudes, there was a main effect of condition [F(1, 28) = 5.86; p = 0.02; g2 = 0.173] showing that the N1 was larger (i.e. more negative) in SCD300 condition (-36.05 lV/m2 ± 3.69) than in the SCD0 condition (-27.91 lV/m2 ± 2.93). There were no other main effects, interactions or latency difference effects evident for the N1 (all F \ 1; p [ 0.5). Summing up the findings on attention-related ERP components, we found that the groups did not display any differential effects across the task conditions and electrodes.
F \ 1.23; p [ 0.5). To examine the interaction of SCD interval 9 group in more detail, we calculated a difference wave between the SCD0 and SCD300 conditions (shown in Fig. 3) and then compared this difference value across groups. The difference between the SCD0 and the SCD300 conditions was larger in younger adults (5.12 ± 0.95) than in older adults (1.23 ± 1.12) [t(28) = -3.69; p = 0.003]. When looking at the topography plots in Fig. 3, it seems that the electrode best reflecting the P3 difference between the SCD0 and the SCD300 condition was Cz for younger adults and FC3 for older adults. However, an analysis revealed that the pattern of results remained the same when quantifying electrode FC3 instead of Cz in the older adults. Using sLORETA, we compared the difference wave of older adults to the difference wave of younger adults in order to examine which brain areas underlie the observed age-related effects. The results indicate that differences between groups were due to activation differences in the TPJ, encompassing the inferior parietal cortex (IPC) (BA 40) and the superior temporal gyrus (STG) (BA39, BA13) (Table 1). In summary, P3 peak amplitudes were differentially modulated across groups and conditions. Activation differences in the TPJ seem to underlie the observed agerelated effects.
P3 The P3 at electrode Cz is shown in Fig. 3. The mixed effects ANOVA revealed no main effect of SCD interval [F(1, 28) = 0.58; p [ 0.4]. There was a main effect of group [F(1, 28) = 16.98; p \ 0.001; g2 = 0.378] showing that the P3 amplitude was smaller in the older adults (17.68 lV/m2 ± 2.43) than in the younger adults (31.03 lV/m2 ± 2.13). Importantly, there was an interaction of SCD interval 9 group [F(1,28) = 7.28; p \ 0.05; g2 = 0.206]. There were no other main effects, interactions or latency difference effects evident for the P3 ERP (all
123
Discussion In the current study, we examined how action cascading processes and the strategy applied to solve multi-demand situations are modulated by aging effects. To this end, we examined performance and neurophysiological processes
Brain Struct Funct
Fig. 3 Event-related current source densities for the P3 in the older adults’ group (left) and the younger adults’ group (right). Time point 0 denotes the time point of STOP stimulus delivery. In the SCD0 condition, the CHANGE stimulus is presented simultaneously while in the SCD300 condition, the onset of the CHANGE stimulus does not occur until 300 ms after the onset of the STOP stimulus. The black and red lines denote the SCD0 and SCD300 conditions, respectively. The blue line denotes the difference wave (SCD0 minus
SCD300). The scalp topography plots show the distribution of the P3 in top of the skull. In the scalp topography plots, cold colors indicate negativity, warm color positivity. The sLORETA plot (middle of the figure) denotes the source obtained from the comparison of the difference waves of the older adults and the younger adults. It shows the difference at the time point of the peak in the difference wave of the younger adults as compared to the older adults. The color scales provides the critical t values
Table 1 Behavioural data in the young and elderly group
limitations of response selection processes (Verbruggen et al. 2008), this likely leads to longer RTs in the SCD0 condition. The neurophysiological data underline this interpretation that limitations at the response selection level underlie the observed age effects on the processing strategy applied during action cascading. The P1 and N1 data did not reveal differential (interaction) effects of age group and any other relevant factor, only main effects of SCD interval. Perceptual and attentional selection processes, as reflected by these ERPs (Herrmann and Knight 2001), are, therefore, not modulated in a way that would explain the behavioral data. This suggests that modulations in perceptual and attentional selection processes are very unlikely to underlie age-related differences in processing strategies during action cascading (Verhaeghen and Cerella 2002; Wasylyshyn et al. 2011). However, the pattern was different for the P3 data. It showed no modulatory effect between the SCD0 and SCD300 conditions in the older adults’ group, while there was a larger P3 in the SCD0 than in the SCD300 condition in the younger adults group. In choice reaction tasks, the P3 reflects an intermediate process between stimulus evaluation and responding (Falkenstein et al. 1994a, b; Verleger et al. 2005) and it has furthermore been shown that the P3 reflects a strategic response selection bottleneck (e.g., Brisson and Jolicoeur 2007; Sigman and Dehaene 2008). The lack of P3 modulation between the SCD conditions in older adults suggests that they may not be able to intensify response selection processes to prioritize one action (i.e., stopping) over the other (i.e., changing). As a consequence, both stopping and changing most likely have
Young group
Elderly group
Go RT
511 ms ± 83
782 ms ± 132
RT SCD0
703 ms ± 90
1063 ms ± 153
RT SCD300
535 ms ± 100
838 ms ± 205
Slope SCD RT function SSRT
-0.56 ± 0.16 244 ms ± 83
-0.75 ± 0.18 565 ms ± 261
The mean ± SD is given
in a stop-change task in older adults and younger adults. We obtained robust effects of aging, as indicated by effect sizes. The behavioral data show that in each of the experimental conditions of the stop-change task, older adults were generally responding more slowly than younger adults (Deline and Juhel 2012; Gazes et al. 2012; Wasylyshyn et al. 2011). Furthermore, there was a greater ageinduced RT increase in the SCD0 condition than in the SCD300 condition, resulting in a steeper slope in older adults than in younger adults. According to the applied framework (cf. Verbruggen et al. 2008), this suggests that older adults employ a more parallel processing strategy than younger adults. It seems that older adults try to processes simultaneously presented stimuli that trigger the execution of different actions more in parallel than younger adults. In other words, they do not prioritize stopping over changing process to the same extent as younger adults, which may reflect known difficulties in elderly to inhibit responses (Cabeza and Dennis 2012). Due to the capacity
123
Brain Struct Funct
to share limited processing capacities. At the behavioral level, the effect of this is reflected by the prolongation of response times in the SCD0 condition. Interestingly, the sLORETA analysis showed that these group-dependent differences in the modulation of the P3 across SCD conditions were due to activation differences in the TPJ (Fjell et al. 2007), including the inferior parietal cortex (IPC) (BA 40) and the superior temporal gyrus (STG) (BA39, BA13). The IPC has been shown to be part of the multiple demand system, which is important for complex multicomponent behavior (Duncan 2010), i.e. in situations where a chaining of different actions is necessary to achieve a task goal. Also, other studies have shown that the TPJ plays a role in the chaining of actions during response selection (e.g., Chersi et al. 2011) and BA40 has been shown to sustain executive control (Collette et al. 2005). Together with the behavioral data, the results of the neurophysiological data on the systems level, therefore, suggest that older adults present with deficits at the response selection level when required to set up a hierarchy of actions. This may lead to an inefficient, more parallel processing strategy of different task goals. A limitation of the study is that the groups were not matched in gender. Concerning the sLORETA results, it is possible that differences in the P3 scalp topography between the groups may underlie effects observed in the temporo-parietal junction. In summary, the study shows that there are age-related differences in the strategy applied during action cascading. Older adults seem to apply a less efficient strategy to cascade different actions. Compared to younger adults, older adults seem to be rather unable to set up a hierarchy of actions, which leads to an inefficient, more parallel processing of different task goals. This deficit is likely due to altered processing of task goals at the response selection level but not due to the processing of stimuli at the attentional selection level. The results suggest that changes in neural processing in the TPJ underlie these age-related modulations in the processing strategy during action cascading. Acknowledgments This work was supported by a Grant from the Deutsche Forschungsgemeinschaft (DFG) BE4045/10-1 and 10-2.
References Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J (1961) An inventory for measuring depression. Arch Gen Psychiatry 4:561–571 Beste C, Saft C (2015) Action selection in a possible model of striatal medium spiny neuron dysfunction: behavioral and EEG data in a patient with benign hereditary chorea. Brain Struct Funct 220:221–228. doi:10.1007/s00429-013-0649-9
123
Beste C, Humphries M, Saft C (2014a) Striatal disorders dissociate mechanisms of enhanced and impaired response selection— evidence from cognitive neurophysiology and computational modelling. Neuroimage Clin 4:623–634. doi:10.1016/j.nicl. 2014.04.003 Beste C, Stock A-K, Epplen JT, Arning L (2014b) On the relevance of the NPY2-receptor variation for modes of action cascading processes. Neuroimage. doi:10.1016/j.neuroimage.2014.08.026 Brisson B, Jolicoeur P (2007) Cross-modal multitasking processing deficits prior to the central bottleneck revealed by event-related potentials. Neuropsychologia 45:3038–3053. doi:10.1016/j.neu ropsychologia.2007.05.022 Buckner RL (2004) Memory and executive function in aging and AD: multiple factors that cause decline and reserve factors that compensate. Neuron 44:195–208. doi:10.1016/j.neuron.2004.09. 006 Cabeza R, Dennis N (2012) Frontal lobes and aging: deterioration and compensation. Princ Front Lobe Funct:628–652. doi:10.1093/ acprof:oso/9780195134971.001.0001 Chersi F, Ferrari PF, Fogassi L (2011) Neuronal chains for actions in the parietal lobe: a computational model. PLoS One 6:e27652. doi:10.1371/journal.pone.0027652 Chmielewski WX, Yildiz A, Beste C (2014) The neural architecture of age-related dual-task interferences. Front Aging Neurosci 6:193. doi:10.3389/fnagi.2014.00193 Collette F, Oliver L, Van der Linden M, Laureys S, Delfiore G, Luxen A, Salmon E (2005) Involvement of both prefrontal and inferior parietal cortex in dual-task performance. Brain Res Cogn Brain Res 24:237–251 Coxon JP, Van Impe A, Wenderoth N, Swinnen SP (2012) Aging and inhibitory control of action: cortico-subthalamic connection strength predicts stopping performance. J Neurosci Off J Soc Neurosci 32:8401–8412. doi:10.1523/JNEUROSCI.6360-11. 2012 Coxon JP, Goble DJ, Leunissen I, Van Impe A, Wenderoth N, Swinnen SP (2014) Functional brain activation associated with inhibitory control deficits in older adults. Cereb Cortex N Y N. doi:10.1093/cercor/bhu165 Deline S, Juhel J (2012) Cognitive modeling of different processing modes in task switching: towards an explanation of the effects of aging on switching cost [WWW Document]. In: Proceedings of ICCM 2012 11th international conference on cognitive modelling. http://act-r.psy.cmu.edu/wordpress/wp-content/uploads/ 2014/02/Deline-2012-ICCM.pdf. Accessed on 13 Jan 2015 Duncan J (2010) The multiple-demand (MD) system of the primate brain: mental programs for intelligent behaviour. Trends Cogn Sci. doi:10.1016/j.tics.2010.01.004 Falkenstein M, Hohnsbein J, Hoormann J (1994a) Effects of choice complexity on different subcomponents of the late positive complex of the event-related potential. Electroencephalogr Clin Neurophysiol 92:148–160 Falkenstein M, Hohnsbein J, Hoormann J (1994b) Time pressure effect on late components of the event-related potential (ERP). J Psychophysiol 8:22–30 Fjell AM, Walhovd KB, Fischl B, Reinvang I (2007) Cognitive function, P3a/P3b brain potentials, and cortical thickness in aging. Hum Brain Mapp 28:1098–1116. doi:10.1002/hbm.20335 Gazes Y, Rakitin BC, Habeck C, Steffener J, Stern Y (2012) Age differences of multivariate network expressions during taskswitching and their associations with behavior. Neuropsychologia 50:3509–3518. doi:10.1016/j.neuropsychologia.2012.09.039 Hartley AA, Jonides J, Sylvester C-YC (2011) Dual-task processing in younger and older adults: similarities and differences revealed by fMRI. Brain Cogn 75:281–291. doi:10.1016/j.bandc.2011.01. 004
Brain Struct Funct Herrmann CS, Knight RT (2001) Mechanisms of human attention: event-related potentials and oscillations. Neurosci Biobehav Rev 25:465–476 Klostermann EC, Braskie MN, Landau SM, O’Neil JP, Jagust WJ (2012) Dopamine and frontostriatal networks in cognitive aging. Neurobiol Aging 33(623):e15–e24. doi:10.1016/j.neurobiola ging.2011.03.002 Miller J, Ulrich R, Rolke B (2009) On the optimality of serial and parallel processing in the psychological refractory period paradigm: effects of the distribution of stimulus onset asynchronies. Cognit Psychol 58:273–310. doi:10.1016/j.cogpsych. 2006.08.003 Mu¨ckschel M, Stock A-K, Beste C (2014) Psychophysiological mechanisms of interindividual differences in goal activation modes during action cascading. Cereb Cortex N Y N 1991(24):2120–2129. doi:10.1093/cercor/bht066 Ness V, Beste C (2013) The role of the striatum in goal activation of cascaded actions. Neuropsychologia 51:2562–2571. doi:10. 1016/j.neuropsychologia.2013.09.032 Nieoullon A, Coquerel A (2003) Dopamine: a key regulator to adapt action, emotion, motivation and cognition. Curr Opin Neurol 16(Suppl 2):S3–S9 Nunez PL, Pilgreen KL (1991) The spline-Laplacian in clinical neurophysiology: a method to improve EEG spatial resolution. J Clin Neurophysiol Off Publ Am Electroencephalogr Soc 8:397–413 Perrin F, Pernier J, Bertrand O, Echallier JF (1989) Spherical splines for scalp potential and current density mapping. Electroencephalogr Clin Neurophysiol 72:184–187 Schwarz W, Ischebeck A (2001) On the interpretation of response time vs onset asynchrony functions: application to dual-task and precue-utilization paradigms. J Math Psychol 45:452–479. doi:10.1006/jmps.2000.1336 Sebastian A, Baldermann C, Feige B, Katzev M, Scheller E, Hellwig B, Lieb K, Weiller C, Tu¨scher O, Klo¨ppel S (2013) Differential effects of age on subcomponents of response inhibition. Neurobiol Aging 34:2183–2193. doi:10.1016/j.neurobiolaging.2013.03.013
Sigman M, Dehaene S (2008) Brain mechanisms of serial and parallel processing during dual-task performance. J Neurosci Off J Soc Neurosci 28:7585–7598. doi:10.1523/JNEUROSCI.0948-08. 2008 Stock A-K, Arning L, Epplen JT, Beste C (2014a) DRD1 and DRD2 genotypes modulate processing modes of goal activation processes during action cascading. J Neurosci Off J Soc Neurosci 34:5335–5341. doi:10.1523/JNEUROSCI.5140-13.2014 Stock A-K, Heintschel von Heinegg E, Ko¨hling H-L, Beste C (2014b) Latent Toxoplasma gondii infection leads to improved action control. Brain Behav Immun 37:103–108. doi:10.1016/j.bbi. 2013.11.004 Verbruggen F, Schneider DW, Logan GD (2008) How to stop and change a response: the role of goal activation in multitasking. J Exp Psychol Hum Percept Perform 34:1212–1228. doi:10. 1037/0096-1523.34.5.1212 Verhaeghen P, Cerella J (2002) Aging, executive control, and attention: a review of meta-analyses. Neurosci Biobehav Rev. doi:10.1016/S0149-7634(02)00071-4 Verhaeghen P, Steitz DW, Sliwinski MJ, Cerella J (2003) Aging and dual-task performance: a meta-analysis. Psychol Aging 18:443–460. doi:10.1037/0882-7974.18.3.443 Verleger R, Jas´kowski P, Wascher E (2005) Evidence for an integrative role of P3b in linking reaction to perception. J Psychophysiol 19:165–181 Wasylyshyn C, Verhaeghen P, Sliwinski MJ (2011) Aging and task switching: a meta-analysis. Psychol Aging 26:15–20. doi:10. 1037/a0020912 Williams BR, Ponesse JS, Schachar RJ, Logan GD, Tannock R (1999) Development of inhibitory control across the life span. Dev Psychol 35:205–213 Yildiz A, Quetscher C, Dharmadhikari S, Chmielewski W, Glaubitz B, Schmidt-Wilcke T, Edden R, Dydak U, Beste C (2014) Feeling safe in the plane: neural mechanisms underlying superior action control in airplane pilot trainees-A combined EEG/MRS study. Hum Brain Mapp. doi:10.1002/hbm.22530
123
